I could solve this issue by making the execution for handlehttprequest to run only on primary node instead of all nodes. Post that I could see that number of threads are not crossing permissible limit and the flow and Nifi services are running fine with out any issues. Not sure, but I think handlehttprequest can't run on a cluster mode with multiple nodes as the endpoint url:port specified will be unique and some TCP sockets are getting locked state and increased the nproc value more than specified value for the nifi user. Thanks, Vish ... View more

Thanks a lot @Matt Burgess for the details on the current limitations and the Jira. I would look at the solution provided by you on the other thread. Appreciate your help on the same. Regards, Vish ... View more

sorry for my ignorance. I could resolve this issue by removing one additional space in " list" while wring the flowfile content generated by groovy script. Now I am able to fetch the data from invoke http by passing flowfile content from executescript. Thanks for the support. Regards, Vish ... View more